4d57e9fc   
   From: info@turneraudio.com.au   
      
   While installing newly potted Hammond SEA1627 OPTs to a pair of Jadis   
   300B mono amps, my mind came onto the idea of using local Single Ended   
   CFB without needing a specially wound SE OPT which has two primary   
   windings, one for anode to B+ and other for cathode to 0V. The SEA1627   
   Hammond has a 40% UL tap so that many pentoads and tetrodes can be set   
   up for 40% UL, or if you like, 60% UL if you connect the single anode   
   winding "around the other way."   
      
   The Jadis has a pair of 300B in SE parallel, and so the UL tap on   
   Hammond OPT is useless unless in future years the owner wishes to   
   switch from 300B to KT120, so PO max could be increased from 17W to   
   30Watts, using the usual UL tap going to screens and the primary   
   having B+ at +430Vdc.   
      
   But what if the UL tap was GROUNDED directly to 0V and ALL the primary   
   placed at oV potential?   
      
   One may wonder how the Fk is that done? The simple answer is to make   
   the B+ supply NON GROUNDED and with its negative sitting on one end of   
   the primary which is 60% of P turns away from the UL tap and have the   
   OP tubes with their cathode bias networks all sitting on the other end   
   of primary which is 40% of P turns away from the UL tap.   
      
   Meanwhile the anode goes to the positive +430Vdc of the B+ supply.   
   The screens may then be taken to a second B+ supply of say +350Vdc,   
   and would optimally be SHUNT regulated.   
      
   The result is that the pentodes or tetroads have internal signal   
   voltage relationships equal to the normal application of 40%   
   Ultralinear taps, but 40% of the Vak is feedback voltage in series   
   with the Vgk input signal.   
      
   Rather than weigh down Google groups with a huge boring talk about the   
   analysis of the result, I give you the result of my calcs indicating   
   the grounded 40%UL tap and floating B+ PSU will reduce THD at near   
   full PO to < 1.5%, and Ra < 400 ohms, and Vg1-0V drive voltage is just   
   under 100Vac where RLa = 2k0 for a single KT120, at least, based on a   
   model  have for 6550EH and KT88EH.   
      
   The problems you will cite are that one needs TWO B+ supplies, and one   
   for OP tubes floats and is prone to noise via stray C on PT. One   
   should use a PT where heaters are wound on as a layer between mains   
   and HT, and heaters are at 0V, so they act as an electrostatic shield.   
   Then one would be wise to have an ES using a turn of foil (without a   
   join to avoid a shorted turn) after the OP tube HT winding.   
      
   The other bother is the drive voltage of 100Vrms. Be brave men, and   
   don't be pussy wussies, and use an EL84 in triode with say 4k7 plus   
   60H choke in series as the feed to +400Vdc with Ia at 14mA. That   
   should give you Ea at about 322Vdc, OK for trioded EL84, and swing   
   will exceed 120Vrms with 2H < 1.5%, which should cancel with 2H of OP   
   stage - rather nicely, you shoud find.   
      
   40% is rather an extreme amount of CFB. 20% is IDEAL, but in this   
   silly world hardly anything made for one thing is useful for something   
   else but nevertheless one can use an SE UL OPT in the manner I have   
   pointed out and if the OPT had UL taps at less than 40% then you are   
   quite fortunate.   
      
   Of course you can use the similar arrangement for Push Pull with two   
   OP tubes and where you'd need to ground the primary CT, and have TWO   
   floating B+ supplies. Consider the OPT with say 40% UL taps, and from   
   one end   
   of primary to other we would have A1, S1, CT, S2, A2 connections for   
   standard UL use for V1 and V2.   
      
   V1 has anode to B+1, then B-1 to A2. V1 cathode is to S1. The other   
   tube, V2 is set up similarly on the other side of PP circuit. Now what   
   could you call this kind of set up? Partial Circlotron.   
      
   Suppose you have the tubes set up for PP and minimal class A. Consider   
   the class B operation. In normal class B the single tube turning on   
   sees RLa = 1/4 RLa-a. But in the case above for PC, the single tube   
   turning on is effectively connected to 70% of the whole primary turns   
   so its class B load becomes 1/2 RLa-a, and this gives the tube a big   
   relief from finding itself connected to a load way too low which has   
   been determined by a fuctard bean counter, and instead the load is   
   twice the ohms. I leave you all to consider the loads actually seen by   
   tubes in class A and AB conditions and maybe you find the PC   
   connection effectively raises the tube loads when compared to plain   
   old UL which has brainwashed so many people into thinking there's no   
   other way.   
      
   BTW, today I tested the Jadis amp after installing the potted Hammond   
   OPT and tried some music. WOW, all I can say is that some bass heavy   
   jazz from Arts FM radio in Canberra sounded just fabulous, and I   
   wanted to EAT THE BASS NOTES. Highs are beatifully clear, without any   
   sign of smear, and the amp just dissapeared to leave the musos in my   
   shed on not-too-good old speakers with not-too-good old EH300B and   
   EH6SN7. The owner has some old RCA 6SN7, and a quad of Emission Labs   
   300Bs. Should sound breathtaking on his much better speakers. Used   
   with mild global NFB, and decent design and better OPT than Jadis   
   provided, the 300Bs seem to make music from heaven.   
      
   Patrick Turner.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)